Numerical methods based strategy and particle filter state estimation for bio process control

Biotechnological processes represent a challenge into the control field, due to its high non linearity. Particularly, continuous alcoholic fermentation by Zymomonas mobilis (Z.m) presents high ethanol performance, but it shows a highly oscillatory behavior in process variables, because of the influence of inhibition dynamics over Biomass, Substrate and Product concentrations; it is a significant challenge from the control perspective. In this work, a new and significant solution for control of biotechnological variables into the fermentation process, based on numerical methods and linear algebra is proposed. In addition, a state estimator of Biomass into the control loop is presented, using to reach this aim, the Non Linear Filtering based on the recursive application of the Bayes rule and Monte Carlo techniques. Specifically, the recursive Bayesian Filters are used. This methodology allows developing the controller design and a tested state estimator for Biomass in Z.m, without the use of complex calculations.